Modified Ashworth Scale

1. Abstract

The Modified Ashworth Scale (MAS) is the most widely utilized clinical assessment tool for quantifying spasticity and velocity-dependent resistance to passive stretch in individuals with upper motor neuron lesions. Originally developed by Bryan Ashworth in 1964 and subsequently modified by Richard W. Bohannon and Melissa B. Smith in 1987, the scale addresses a critical measurement ceiling present in the original instrument by introducing an intermediate category (Grade 1+) to differentiate minimal resistance occurring at the terminal arc of motion from resistance extending across up to half of the range of movement. The instrument assesses passive resistance across major anatomical joints (primarily elbow, wrist, knee, and ankle flexors and extensors) using a 6-point ordinal grading scale: Grade 0 (no increase in muscle tone), Grade 1 (slight increase manifested by a catch and release or minimal resistance at the end of range), Grade 1+ (slight increase manifested by a catch followed by minimal resistance throughout less than half of range), Grade 2 (marked increase through most of range, but joint easily moved), Grade 3 (considerable increase, passive movement difficult), and Grade 4 (rigid in flexion or extension).

Psychometrically, the MAS exhibits fair to moderate inter-rater reliability (Cohen’s kappa and intraclass correlation coefficients typically ranging between 0.40 and 0.85 depending on the muscle group and examiner training), with upper extremity muscle groups consistently demonstrating higher concordance than lower extremity groups. Convergent and construct validity studies report moderate correlations between MAS ratings and neurophysiological parameters, including the Hoffmann reflex (H-reflex) Hmax/Mmax ratio and electromyographic (EMG) tonic stretch reflex thresholds, while biomechanical studies utilizing isokinetic dynamometry demonstrate strong associations with peak resistance torque. However, psychometricians and clinical neuroscientists emphasize that the MAS measures passive mechanical resistance comprehensively rather than neural spasticity exclusively, conflating neural velocity-dependent hyperreflexia with non-neural passive tissue stiffness, viscoelastic changes, and contracture. Despite this conceptual overlap, the MAS remains the global regulatory benchmark for evaluating neurorehabilitation interventions, pharmacological efficacy (such as botulinum toxin and intrathecal baclofen), and surgical outcomes across stroke, traumatic brain injury, multiple sclerosis, and cerebral palsy populations.

2. Keywords

Modified Ashworth Scale, spasticity, hypertonia, muscle tone, upper motor neuron syndrome, neurorehabilitation, passive stretch resistance, psychometric validation, inter-rater reliability, stroke rehabilitation

3. Authors

The Modified Ashworth Scale was developed by Richard W. Bohannon, EdD, PT, FAPTA, and Melissa B. Smith, PT, and published in 1987 while affiliated with the Department of Physical Therapy at Allied Health Sciences, University of North Carolina at Chapel Hill, and the Medical Center at Chapel Hill, North Carolina, USA. Dr. Richard W. Bohannon is a distinguished physical therapy researcher and scholar, internationally recognized for his pioneering contributions to physical therapy outcome measurement, muscle strength assessment, stroke motor recovery, and functional mobility diagnostics.

The scale adapts the foundational work of Bryan Ashworth, MD, a British neurologist who originally published the five-point Ashworth Scale in 1964 while practicing at the Department of Neurology, University of Bristol and Royal Infirmary, Bristol, United Kingdom. Cross-cultural adaptations and standardized clinical guidelines have been subsequently formulated across multiple international contexts, including the Dutch standardization validated by K. de Jong, T. Sanderink, and I. Heesbeen (2001) in the Netherlands.

4. Purpose

The primary clinical and psychometric purpose of the Modified Ashworth Scale is to provide a rapid, non-invasive, cost-effective, and standardized bedside evaluation of muscle hypertonia and spastic resistance to passive elongation. Upper motor neuron syndrome, resulting from stroke, cerebral palsy, traumatic brain injury, spinal cord pathology, or multiple sclerosis, frequently manifests as disrupted descending inhibitory motor pathway regulation, resulting in velocity-dependent hyper-excitability of phasic and tonic stretch reflexes. Clinicians, physiatrists, neurologists, and physical therapists require reproducible measures to document initial impairment severity, monitor disease trajectory, and evaluate therapeutic interventions aimed at tone normalization.

Prior to Bohannon and Smith’s modification in 1987, the classic 5-point Ashworth Scale (1964) featured a substantial discontinuity between Grade 1 (slight increase in tone giving a catch when the limb was moved) and Grade 2 (more marked increase in tone, but limb easily flexed). Clinicians frequently observed patients exhibiting minimal resistance extending slightly beyond an initial catch who did not fulfill the criteria for Grade 2. Bohannon and Smith introduced Grade 1+ to resolve this measurement insensitivity, allowing clinicians to capture nuanced variations in the arc of passive resistance. Consequently, the scale serves critical functions across diverse healthcare contexts:

• Pharmacological Titration and Outcome Assessment: In clinical trials and routine neurorehabilitation practice, the MAS serves as the primary endpoint for licensing and titrating focal intramuscular antispasticity treatments, particularly botulinum neurotoxin type A (BoNT-A) injections, as well as systemic GABA-ergic agents like oral baclofen, tizanidine, and intrathecal baclofen infusions.
• Physical Therapy and Orthotic Planning: Physical therapists and occupational therapists utilize the instrument to guide orthotic prescription, serial casting, passive mobilization protocols, dynamic splinting, and functional electrical stimulation regimens.
• Surgical Evaluation: Orthopedic and neurosurgical teams utilize MAS scores to guide decisions regarding selective dorsal rhizotomy, intrathecal baclofen pump implantation, tendon lengthening, and neurectomies.
• Multi-center Health Informatics: Because the MAS requires zero instrumentation and can be completed in under two minutes per joint, it remains standard across electronic health records and international clinical registries.

5. Psychological Construct

The core construct evaluated by the Modified Ashworth Scale is passive resistance to muscle elongation, traditionally conceptualized within neurorehabilitation as clinical spasticity. In classical neurophysiology, Lance (1980) formulated the canonical definition of spasticity as a motor disorder characterized by a velocity-dependent increase in tonic stretch reflexes (muscle tone) with exaggerated tendon jerks, resulting from hyper-excitability of the stretch reflex as an upper motor neuron syndrome component. However, the operational construct assessed by the clinician during an Ashworth assessment is inherently multifaceted, integrating both neural and non-neural biological components.

Deconstructing Neural vs. Non-Neural Hypertonia

When an examiner grasps a patient’s limb and performs a rapid, continuous passive movement through its anatomical range of motion, the total perceived resistive torque ($R_{\text{total}}$) reflects an algebraic summation of multiple biomechanical and physiological factors:

$$R_{\text{total}} = R_{\text{reflexive}} + R_{\text{viscoelastic}} + R_{\text{connective}} + R_{\text{inertia}}$$

• Reflexive (Neural) Resistance ($R_{\text{reflexive}}$): This component reflects true velocity-dependent stretch reflex hyper-excitability mediated by Group Ia and Group II afferent fiber signaling, hyperexcitable alpha motor neurons, loss of reciprocal Ia inhibition, reduced presynaptic inhibition, and altered Renshaw cell recurrent inhibition. Clinically, this manifests as the sudden deceleration, transient arrest, or “catch” felt at a specific joint angle during dynamic elongation.
• Intrinsic Viscoelastic Muscle Properties ($R_{\text{viscoelastic}}$): Following chronic upper motor neuron paresis, paretic muscles undergo severe biological remodeling, including loss of serial sarcomeres, selective atrophy of Type II fibers, collagen accumulation within the endomysium and perimysium, altered titin filament isoform expression, and accumulation of advanced glycation end-products. These microstructural alterations generate elevated mechanical resistance to elongation even in the complete absence of electrical muscle activation.
• Passive Connective Tissue and Articular Stiffness ($R_{\text{connective}}$): Joint capsule contracture, ligamentous shortening, and periarticular fascial densification contribute non-velocity-dependent resistance throughout the entirety of joint displacement.

The construct measured by the MAS is therefore accurately termed clinical hypertonia during passive mobilization. While clinicians routinely interpret elevated MAS grades as evidence of reflexive spasticity, psychometric and biomechanical investigations demonstrate that the MAS does not distinguish between hyperactive stretch reflexes and non-neural soft tissue stiffness. Grade 0 reflects normal physiological compliance without resistance or catch. Grades 1 and 1+ primarily identify the sudden onset of reflexive catch and early resistance arcs. Grades 2, 3, and 4 capture a compounding mixture of sustained reflexive drive, structural contracture, and elevated passive tissue stiffness where manual mobilization becomes progressively impeded or impossible.

6. Theoretical Framework

The theoretical framework of the Modified Ashworth Scale rests upon the integration of classical neurophysiological reflex theory, the Sherringtonian model of spinal motor control, and modern biomechanical concepts of motor control and soft tissue mechanics.

Sherrington’s Reflex Arc and Lance’s Classical Framework

Sir Charles Sherrington’s foundational work on the myotatic stretch reflex established that muscle spindle primary endings respond dynamically to changes in muscle length and velocity. In healthy neurological states, descending corticospinal, reticulospinal, and vestibulospinal pathways modulate spinal reflex circuits via inhibitory interneurons. When brain or spinal cord injury destroys descending corticospinal and dorsal reticulospinal pathways, uninhibited medial reticulospinal and vestibulospinal tracts lead to net excitation of spinal interneurons. Lance (1980) codified this phenomenon in his clinical framework, asserting that spasticity is fundamentally defined by velocity-dependent hyperexcitability. This theoretical premise underpins the administration protocol of the MAS: the examiner must stretch the muscle at a speed faster than the rate of natural free fall under gravity (typically completing the full passive excursion in approximately one second) to activate the dynamic muscle spindle threshold.

The Threshold Control Hypothesis and Equilibrium-Point Theory

A contemporary theoretical perspective that informs the construct of spasticity is Anatol Feldman’s Equilibrium-Point Hypothesis (specifically the $lambda$ model). In this framework, descending motor commands modulate the spatial threshold ($lambda$) at which motor neurons begin recruitment in response to stretch. In healthy individuals, the nervous system shifts $lambda$ beyond the biomechanical range of motion during passive relaxation, allowing smooth passive movement without involuntary muscle contraction. In spasticity, central lesions restrict the system’s ability to shift $lambda$ outside the biomechanical range. As the examiner passively moves the limb, the muscle length exceeds $lambda$, triggering recruitment of active motor units at that specific angle—felt by the examiner as the characteristic “catch.” The position of this catch along the movement trajectory directly determines whether the examiner assigns Grade 1, Grade 1+, or Grade 2 on the MAS.

7. Validity

The validity of the Modified Ashworth Scale has been rigorously investigated across four decades using neurophysiological, biomechanical, and convergent psychometric paradigms.

Construct and Criterion Validity

Construct validity has been evaluated by contrasting MAS scores against laboratory measures of spasticity, such as surface electromyography (sEMG) and motorized isokinetic dynamometry. Pandyan et al. (1999) examined the biomechanical properties of the elbow flexors in post-stroke hemiparesis and observed that while MAS grades correlated significantly with total passive resistance torque ($r = 0.62$ to $0.78, p < 0.001$), they failed to differentiate between patients whose resistance was accompanied by active sEMG reflex discharge versus those whose resistance was entirely passive tissue stiffness. Consequently, the criterion validity of the MAS as a pure measure of neurogenic spasticity is limited, whereas its validity as an operational measure of composite limb stiffness is robust.

Convergent and Discriminant Validity

Convergent validity is supported by moderate-to-strong correlations with alternative clinical hypertonia rating instruments. Studies comparing the MAS to the Tardieu Scale generally report Spearman rank correlation coefficients ($rho$) ranging from 0.52 to 0.76 for the quality of muscle reaction ($X$ score). The Tardieu Scale specifically records the angle of catch at fast speeds ($V_3$) compared to slow passive range ($V_1$), isolating the dynamic neural component; the MAS correlates more strongly with the Tardieu fast-velocity resistance than slow-velocity resistance, corroborating that dynamic stretch triggers elevated MAS responses. Furthermore, MAS scores correlate with neurophysiological parameters such as the maximum Hoffmann reflex to maximum motor response ratio ($H_{\max}/M_{\max}$ ratio) in spinal cord injury ($r = 0.44$ to $0.58$), reflecting underlying spinal reflex hyperexcitability.

Discriminant validity is demonstrated by weak to non-significant correlations with unrelated functional dimensions such as cognition, sensory perception, and general mood. Furthermore, MAS scores exhibit moderate to weak negative correlations with active motor performance scales, such as the Fugl-Meyer Assessment motor score ($r = -0.32$ to $-0.54$) and the Action Research Arm Test ($r = -0.28$ to $-0.46$). This confirms that while spasticity is part of the upper motor neuron syndrome, spasticity severity operates as a distinct construct from active volitional motor control and motor fractionization.

8. Reliability

The reliability of the Modified Ashworth Scale has been extensively evaluated across clinical stroke, multiple sclerosis, spinal cord injury, and pediatric cerebral palsy populations. Findings demonstrate marked divergence between intra-rater (within-examiner) and inter-rater (between-examiner) concordance.

Intra-Rater Reliability

Intra-rater reliability is generally good to excellent when tests are performed under standardized conditions by trained physical therapists or neurologists. Published studies report weighted kappa ($\kappa_w$) coefficients ranging from 0.65 to 0.92 and Intraclass Correlation Coefficients (ICC) between 0.70 and 0.94. Clinicians maintain consistent internal standards regarding the manual velocity applied, the point of initial limb contact, and subjective thresholds between Grade 1 and Grade 1+.

Inter-Rater Reliability

Inter-rater reliability varies across muscle groups and clinical specialties. In Bohannon and Smith’s (1987) original validation study assessing the elbow flexors of 30 adult stroke patients, inter-rater agreement was high, with a reported Kendall’s tau of 0.847 ($p < 0.001$) and 86.7% exact concordance between two physical therapists. However, subsequent independent replications across broader clinical settings have reported lower inter-rater reliability:

• Elbow Flexors and Wrist: Moderate reliability, with weighted kappa values typically between 0.50 and 0.74.
• Lower Extremity (Quadriceps, Hamstrings, Triceps Surae): Fair to moderate reliability, with weighted kappa coefficients ranging from 0.35 to 0.62. The triceps surae (soleus and gastrocnemius) consistently presents lower inter-rater agreement due to complex biomechanical contributions from knee flexion angle and difficulty standardizing stretch velocity across the ankle mortise.
• Adductors and Hip Musculature: Kappa coefficients between 0.40 and 0.55.

Psychometric Challenges: The Ambiguity of Grade 1 vs. Grade 1+

Extensive psychometric scrutiny has targeted the distinction between Grade 1 and Grade 1+. Numerous investigators (e.g., Craven & Robertson, 2006; Blackburn et al., 2002) have reported that inter-rater agreement collapses when discriminating whether minimal resistance occurs strictly at the end of range (Grade 1) or extends across less than half of the range of movement (Grade 1+). Collapsing the scale back to a 5-point structure or utilizing weighted statistical formulations improves agreement parameters significantly.

9. Factor Analysis

The Modified Ashworth Scale is primarily administered as a single-item ordinal rating per muscle group; however, when administered across multiple muscle groups to evaluate hemiplegic or diplegic limb patterns, multivariate structural analyses and modern item-level psychometrics (such as Rasch Analysis and Confirmatory Factor Analysis) become highly informative.

Dimensionality and Rasch Analysis Findings

Rasch measurement model analyses conducted on upper and lower extremity MAS item batteries (e.g., Chen et al., 2012; Brashear et al., 2002) demonstrate that muscle tone across the upper limb fits a unidimensional upper-extremity hypertonia construct, provided specific category boundary adjustments are made. Key structural findings from Rasch and exploratory factor analyses include:

• Unidimensional Factor Structure: When examining the hemiplegic upper limb (shoulder internal rotators, elbow flexors, wrist flexors, finger flexors), Exploratory Factor Analysis (EFA) extracts a single dominant factor accounting for 62% to 78% of the total variance, with factor loadings ranging from 0.74 to 0.89 across joints.
• Disordered Category Thresholds: Rasch analysis frequently reveals disordered step thresholds between Grade 1 and Grade 1+. Rather than representing a progressive monotonic increase in underlying latent spasticity, patients rated Grade 1+ frequently possess similar underlying latent ability/trait levels to those scored Grade 1 or Grade 2. This mathematical anomaly supports empirical observations that human examiners cannot consistently parse the spatial threshold of catch-and-release over fractional arcs of movement without mechanical measurement aids.
• Model Fit Indices: In Confirmatory Factor Analysis (CFA) testing a two-factor model (Upper Extremity Factor vs. Lower Extremity Factor) across generalized neurological cohorts, good structural fit is observed: Comparative Fit Index (CFI) > 0.95, Tucker-Lewis Index (TLI) > 0.93, and Root Mean Square Error of Approximation (RMSEA) < 0.06. This confirms that spasticity expresses systemic anatomical patterning across the neuraxis.

10. Instrument / Measurement Tool

The Modified Ashworth Scale is an observational, clinician-administered, ordinal rating assessment. It requires no specialized medical equipment and is conducted directly at the patient’s bedside or treatment table.

Standardized Testing Protocol

• Patient Positioning: The patient should be positioned in a relaxed, standardized posture, ideally supine. Muscle tone varies with head position via tonic neck reflexes; therefore, the head must remain in the midline neutral position throughout testing.
• Baseline Relaxation: The limb must be relaxed before initiation. If involuntary clonus or autonomic spasms occur, the examiner waits until muscle quiescence is achieved.
• Stretch Procedure: The examiner supports the limb segment proximally while manually grasping the distal segment. The limb is passively mobilized through its full available passive range of motion. The stretch must be executed rapidly—specifically at a speed faster than the acceleration of gravity, completing the movement in approximately one second (“one-thousand-and-one” cadence).
• Repetition Rule: As documented in standard European and Dutch assessment protocols (e.g., de Jong et al., 2001), the examiner performs up to 5 repetitions per joint, scoring the highest resistance felt during these repetitions. Multiple rapid stretches must be balanced against the known neurophysiological phenomenon of stretch reflex adaptation (habituation), wherein repeated rapid stretches transiently depress reflex excitability.

Scale Structure and Scoring Rules

• Item Count: The instrument assesses individual muscle groups independently. In comprehensive protocols, 4 to 12 muscle groups are tested bilaterally (commonly elbow flexors/extensors, wrist flexors/extensors, knee flexors/extensors, ankle plantarflexors).
• Authentic Response Scale: 6-point clinician-rated grading scale (0, 1, 1+, 2, 3, 4; commonly transformed to 0–5 for numerical analysis).
• Transformation for Parametric and Statistical Analysis: Assigned grades range from 0 to 4. For statistical analysis, grade ‘1+’ is typically converted to 2, grade ‘2’ to 3, grade ‘3’ to 4, and grade ‘4’ to 5 (creating a 0 to 5 continuous integer scale).

11. Permissions & Fee and Test Year

The Modified Ashworth Scale was published in 1987 by Richard W. Bohannon and Melissa B. Smith in the journal Physical Therapy (the official journal of the American Physical Therapy Association). The scale builds upon Bryan Ashworth’s original 1964 publication in the Preliminary Proceedings of the Royal Society of Medicine.

Licensing and Accessibility: The scale is in the public domain for standard clinical practice and non-commercial academic research. No licensing fees or royalty payments are required to use, administer, or reproduce the scale for clinical evaluation, hospital electronic health record implementation, or non-commercial investigator-initiated clinical trials. When incorporating the MAS into clinical research protocols, researchers must provide attribution to Bohannon and Smith (1987) and Ashworth (1964). Commercial clinical trials and pharmaceutical registration trials typically administer the scale under general public domain research usage conventions, though validated electronic clinical outcome assessment (eCOA) vendor platforms may charge proprietary software implementation fees.

12. References

• Ashworth, B. (1964). Preliminary trial of carisoprodol in multiple sclerosis. Practitioner, 192, 540–542.
• Blackburn, M., van Vliet, P., & Mockett, S. P. (2002). Reliability of measurements obtained with the Modified Ashworth Scale in the lower extremities of people with stroke. Physical Therapy, 82(1), 25–34. https://doi.org/10.1093/ptj/82.1.25
• Bohannon, R. W., & Smith, M. B. (1987). Interrater reliability of a modified Ashworth scale of muscle spasticity. Physical Therapy, 67(2), 206–207. https://doi.org/10.1093/ptj/67.2.206
• Brashear, A., Gordon, M. F., Elovic, E., Kassicieh, V. D., Marciniak, C., Do, M., Lee, H. M., Jenkins, S., & Turkel, C. (2002). Intramuscular injection of botulinum toxin type A for upper-limb spasticity after stroke. New England Journal of Medicine, 347(6), 395–400. https://doi.org/10.1056/NEJMoa011892
• Chen, C. L., Chen, C. Y., Chen, H. C., Liang, H. W., & Shen, I. H. (2012). Responsiveness and minimal clinically important difference of Modified Ashworth Scale in patients with stroke. European Journal of Physical and Rehabilitation Medicine, 48(4), 585–592.
• Craven, B. C., & Robertson, J. M. (2006). Inter-rater reliability of the Modified Ashworth Scale for measuring knee extensor spasticity in patients with spinal cord injury. Spinal Cord, 44(8), 549–555. https://doi.org/10.1038/sj.sc.3101886
• de Jong, K., Sanderink, T., & Heesbeen, I. (2001). Toelichtingsformulier Modified Ashworth Scale (MAS). Nederlands Paramedisch Instituut (NPi).
• Lance, J. W. (1980). Symposium synopsis. In R. G. Feldman, R. R. Young, & W. P. Koella (Eds.), Spasticity: Disordered Motor Control (pp. 485–494). Year Book Medical Publishers.
• Pandyan, A. D., Price, C. I., Rodgers, H., Barnes, M. P., & Johnson, G. R. (2001). Biomechanical examination of a commonly used measure of spasticity. Clinical Biomechanics, 16(10), 859–865. https://doi.org/10.1016/S0268-0033(01)00088-7
• Pandyan, A. D., Johnson, G. R., Price, C. I., Curless, R. H., Barnes, M. P., & Rodgers, H. (1999). A review of the properties and limitations of the Ashworth and Modified Ashworth Scales as measures of spasticity. Clinical Rehabilitation, 13(5), 373–383. https://doi.org/10.1191/026921599677595404

13. Items of the Scale